Search

This study utilized a multiple-input multiple-output synthetic aperture radar (MIMO-SAR) technology employing a frequency-modulated continuous wave (FMCW) millimeter-wave MIMO radar to execute two-dimensional (2D) and three-dimensional (3D) imaging of objects. By analyzing the intermediate frequency (IF) signal generated from the transmitted and received chirp signals, Range FFT spectra with information on the distance to the object and on the reflection intensity according to frequency were obtained. The integration of reflection intensity provided a comprehensive image at varying distances, while the 2D cross-sectional image of an object inside a concealed cardboard achieved a spatial resolution of less than 1 mm. Through innovative data processing techniques, such as the spatial translation filter, we succeeded in the reconstruction of a monolayered 3D image of a clothed human body’s surface. Moreover, the system demonstrated its potential for security checks by effectively identifying dangerous items concealed by human.

In this study, we report high-performance amorphous Ga2O3 metal-oxide (AMO) thin film transistor (TFT) using an low-temperature solution-process coupling with In alloy engineering. In doping can lower the activation temperature of gallium oxide and increase the oxygen vacancy concentration to further activate the device. The optical bandgap of IGO film can be changed from 5.3 to 4.25 eV with the In doping concentration ( $\text{C}_{\mathrm{ In}}$ ) increasing from 0 % to 50 %. All TFTs with IGO channels exhibit n-type transistor characteristics and the evolution of their key electrical parameters with the In-dopant is well elucidated by the structural and morphological characterization. With the increase of $\text{C}_{\mathrm{ In}}$ , the performance of the device becomes better. Finally, a saturation field-effect mobility of 3.63 cm2V−1s−1, a current on/off ratio of 106, and a threshold voltage of 2.5 V are achieved by the In0.5Ga0.5O ( $\text{C}_{\mathrm{ In}}= 50$ %) based device. The In0.5Ga0.5O TFT also demonstrates good bias stress stability. Under the action of 20 V and −20 V gate bias for 3000 s, the ${\Delta }\text{V}_{\mathrm{ TH}}$ is +2.27 V and −1.95 V, respectively.

Quantum image processing will be an important issue in the age of quantum computers. Quantum image processing generally includes quantum image representations, quantum image processing algorithms, and quantum image measurement. Among them, quantum image representation is a very important part of quantum image processing. Therefore, it is of great significance for quantum image processing to study the representation of quantum images. First, we elaborate on the main quantum image representations and analyze, compare and summarize them. This paper focuses on the definition of these quantum image representations the preparation of quantum images and the number of quantum bits required as well as the computational complexity. Then, this paper analyzes, compares and summarizes the similarities and differences between these quantum image representations. Finally, the challenges and future development in the field of quantum image processing are summarized and forecasted. This work will help researchers to understand the advances related to quantum image representations in the field of quantum image processing and is therefore of some reference value.

Community detection is an important research direction in complex network analysis that aims to detect the community structure in networks via clustering operation, which has an important application value and practical significance in mining potential network information. Genetic algorithm (GA) is commonly used in community detection to solve NP-hard problems caused by modularity optimization effectively. However, in terms of GA, the problems of global search performance and slow convergence still remain due to the low accuracy of initial population generation and the randomness of the mutation operator. Thus, a novel local structural similarity-based GA (LSSGA) is proposed in this study to address such problems. It mainly contains two innovations: First, a new generation strategy for initial population based on local structural similarity and roulette wheel selection is designed to improve the quality of initial individuals, at the same time, maintain the diversity of the initial population. Secondly, an effective mutation operator based on label propagation and local structural similarity is proposed, which can quickly achieve targeted and effective mutation. The performance of LSSGA has been verified in many detailed aspects such as the convergence of initialization strategy, the effectiveness of overcome falling into local optimum, and rationality of objective function. Experimental results of both synthetic and real-world networks demonstrate the effective performance of the proposed method via systematic comparison with other existing advanced algorithms.

Volume rendering is an emerging technique widely used in the medical field to visualize human organs using tomography image slices. In volume rendering, sliced medical images are transformed into attributes, such as color and opacity through transfer function. Thus, the design of the transfer function directly affects the result of medical images visualization. A well-designed transfer function can improve both the image quality and visualization speed. In one of our previous paper, we designed a multi-dimensional transfer function based on region growth to determine the transparency of a voxel, where both gray threshold and gray change threshold are used to calculate the transparency. In this paper, a new approach of the transfer function is proposed based on clustering analysis of gray-gradient mode histogram, where volume data is represented in a two-dimensional histogram. Clustering analysis is carried out based on the spatial information of volume data in the histogram, and the transfer function is automatically generated by means of clustering analysis of the spatial information. The dataset of human thoracic is used in our experiment to evaluate the performance of volume rendering using the proposed transfer function. By comparing with the original transfer function implemented in two popularly used volume rendering systems, visualization toolkit (VTK) and RadiAnt DICOM Viewer, the effectiveness and performance of the proposed transfer function are demonstrated in terms of the rendering efficiency and image quality, where more accurate and clearer features are presented rather than a blur red area. Furthermore, the complex operations on the two-dimensional histogram are avoided in our proposed approach and more detailed information can be seen from our final visualized image.

With the popularity of surveillance cameras and the development of deep learning, significant progress has been made in the field of smart surveillance. Face recognition is one of the most important yet challenging tasks in human-centered smart surveillance, especially in public security, criminal investigation and anti-terrorism, and so on. Although, the state-of-the-art algorithms for face recognition have achieved dramatically improved results and have been widely applied in authentication scenario, the occlusion problem on face is still one of the critical issues for personal identification in smart surveillance, especially in the occasion of terrorist searching and identification. To address this issue, this paper proposed a new approach for eyes-to-face synthesis and personal identification for human-centered smart surveillance. An end-toend network based on conditional generative adversarial networks (GAN) is designed to generate the face information based only on the available data of eyes region. To obtain photorealistic faces and identitypreserving information, a synthesis loss function based on feature loss, GAN loss, and total variation loss is proposed to guide the training process. Both the subject and objective experimental results demonstrated that the proposed method can preserve the identity based on eyes-only data, and provide a potential solution for the identification of person even in the case of face occlusion.

The wider use of wearable devices for electroencephalogram (EEG) data capturing provides a very useful way for the monitoring and self-management of human health. However, the large volumes of data with high dimensions cause computational complexity in EEG data processing and pose a great challenge to the use of wearable EEG devices in healthcare. This paper proposes a new approach to extract the structural information of EEG data and tackle the curse of dimensionality of the EEG data. A set of methods for dimensionality reduction (DR)-like linear discriminant analysis (LDA) and their improved methods have been developed for EEG processing in the literature. However, the existing LDA-related methods suffer from the singularity problem or expensive computational cost, and none of existing methods take into consideration the structure of the projection matrix, which is crucial for the extraction of the structural information of the EEG data. In this paper, a new method called a regularized matrix discriminant analysis (R-MDA) is proposed for EEG feature representation and DR. In the R-MDA, the EEG data are represented as a data matrix, and projection vectors are reshaped to be a set of projection matrices stacking together. By reformulating the LDA as a least-square formulation and imposing specified constraint on each projection matrix, the new R-MDA has been constructed to effectively reduce EEG dimensions and capturing the structural information of the EEG data. Experimental results demonstrate that this new R-MDA outperforms the existing LDA-related methods, including achieving improved accuracy with significant DR of the EEG data. This offers an effective way to enable wearable EEG devices be applicable in human-centered health monitoring.

A time-delay estimation method is proposed for single-input single-output (SISO) delay systems. A copula is constructed from uniformly sampled input-output data to build the relationship between the input and output of the system. Taking time-delay as the parameter to be estimated, the logarithm likelihood function is formulated and separated into two parts, i.e., the dependence structure likelihood and the margin-related part. Then, the time-delay is obtained through maximizing the dependence structure likelihood. The highlight of this method is that the structure prior to the studied system is not required to estimate the time-delay. Comparative simulation examples are provided to demonstrate the validity of this method.

The variability in the propagation pathway in epilepsy is a main factor contributing to surgical treatment failure. Ways to accurately capture the brain propagation network and quantitatively assess its evolution remain poorly described. This work aims to develop a dynamic step effective network (dSTE) to obtain the propagation path network of multiple seizures in the same patient and explore the degree of dissimilarity. Multichannel stereo-electroencephalography (sEEG) signals were acquired with ictal processes involving continuous changes in information propagation. We utilized high-order dynamic brain networks to obtain propagation networks through different levels of linking steps. We proposed a dissimilarity index based on singular value decomposition to quantitatively compare seizure pathways. Simulated data were generated through The Virtual Brain, and the reliability of this method was verified through ablation experiments. By applying the proposed method to two datasets consisting of 29 patients total, the evolution processes of each patient’s seizure networks was obtained, and the within-patient dissimilarities were quantitatively compared. Finally, three types of brain network connectivity patterns were found. Type I patients have a good prognosis, while type III patients are prone to postoperative recurrence. This method captures the evolution of seizure propagation networks and assesses their dissimilarity more reliably than existing methods, demonstrating good robustness for studying the propagation path differences for multiple seizures in epilepsy patients. The three different patterns will be important considerations when planning epilepsy surgery under sEEG guidance.

The coral reef geomorphic zone classification (CRGZC) map can provide a wealth of information for coastal management and protection. Remote sensing plays an important role in CRGZC by virtue of its speed, wide range, and low cost. Although many excellent results have been achieved in this field, there are still some shortcomings. With the development of machine learning, such methods are gradually introduced to CRGZC, yet the research and application of deep learning methods are still relatively few. In this article, based on ICESat-2 data and multispectral images, a deep learning model coupled with convolutional neural network (CNN) and random forest (RF) was proposed for coral reef geomorphic zone classification (CR_CRGZC). First, the priori bathymetry points were extracted from ICESat-2. Then, a near-shore bathymetry map was generated using a log-ratio model. Finally, topographic data and multispectral images were combined to achieve CRGZC through CR_CRGZC. The northeastern part of Coffin Island (CI) and the southern part of Punta Vaquero (PV) in Puerto Rico Island were selected as study areas. By comparing the classification results with those of CNN, RF, and maximum likelihood classification, CR_CRGZC outperformed the other classification methods. By quantitative analysis, the OA and Kappa coefficients of CR_CRGZC were 91.91% and 0.9013 in the CI region; and 89.91% and 0.8735 in the PV region, respectively. Under the same environmental requirements, this approach can map high-precision submeter CRGZC maps, providing a database for dynamic coral reef habitat mapping, which contributes to marine coastal ecosystem protection and coastal underwater topography monitoring.

Deep learning has been utilized for hyperspectral image (HSI) classification in recent years, with notable performance improvements. In particular, convolutional neural networks (CNNs) methods have achieved major advancements in this area. However, there are some drawbacks to the existing CNN-based HSI classification approaches: 1) the lack of effective and simple feature representations in CNNs, which overlook the effects of spectral differences and spatial contextual information; 2) the classification model has an enormous network complexity as a result of its numerous training parameters and high computational requirements; and 3) the category samples in HSI data exhibit a significant long tail distribution issue, which affects the classification performance of HSI. To address these issues, we propose a lightweight spectral–spatial neural architecture with multiattention feature extraction (LSSMA) for HSI classification. The main work consists of three areas: 1) A spectral feature extraction and fusing module is created to facilitate the fusing of spectral–spatial features while reducing the number of trainable parameters and computational complexity of the model. This module uses convolutions of various kernel sizes for residual connection and feature fusion, and introduces group convolution to achieve efficient feature representation. 2) To utilize the spectral–spatial correlation of HSI data to its fullest, a multiscale convolutional activation guided attention mechanism is designed and the position attention module is referenced, which can capture spectral differences and spatial contextual relationships between ground objects. and 3) Focal loss is applied in computer vision tasks to the LSSMA model to enhance its capacity to handle category imbalance. Experimental results on four publicly available hyperspectral datasets show that the method obtains better classification performance at a lower computational cost.

Industrial robots are extensively utilized in handling, assembly, and welding tasks owing to their expansive workspace, scalability, flexibility, and cost-effectiveness. However, their inadequate absolute positioning accuracy significantly impedes their application in precise operational scenarios. To enhance robot positioning accuracy, the hysteresis error induced by gear meshing backlash is considered. Firstly, the impact of joint hysteresis on robot positioning errors is analyzed, the notion of modified joint space is introduced, and the similarity theory of error in modified joint space is analyzed. Secondly, for the problem of parameter overfitting of the universal Kriging model, a method of dynamically determining the basis function set by using the genetic algorithm is proposed. Finally, the target trajectory is corrected by a feed-forward iterative compensation algorithm. An experiment on a tandem industrial robot SMART5 NJ 220-2.7 is conducted to demonstrate the effectiveness of the compensation. The experimental results show that the error caused by joint hysteresis is significant, with joint 1 notably affecting y axis positioning accuracy, while joints 2 and 3 predominantly influence x axis positioning accuracy. Furthermore, cross-validation tests verified the good anti-overfitting effect of optimized Kriging for models with multiple input parameters and the good fitting accuracy of the modified space model for hysteresis errors. Moreover, after employing MJS&GPS+GA error modeling and feed-forward iteration compensation, the average absolute positioning error of the trajectory decreased by 81% to 0.09252 mm, and the maximum absolute positioning error decreased by 59% to 0.27713 mm.

In order to improve reliability of distributions, the neutrals are usually insulated or compensated with Peterson coil. Thus, the currents of single-phase to ground fault are weak, leading to difficulty in fault line selection problem. This paper investigates an evidence theory based integrated fault line selection method which employs multiple fault components to improve the performances of selection for neutral compensated distributions. Aim of this method is through the fusion strategy of the multi-fold unreliable fault components to arrive at more reliable solution. Three kinds of fault components including the transient, the abrupt changing, and the damped DC are developed. Fault measures are introduced for each line to quantitatively describe the likelihood of the line being the fault line. The fault components are extracted by dual-tree complex wavelet and atom decomposition. And the algorithms of calculating fault measures for the three components are developed according to the principle of the fault. The basic probability assignment (BPA) functions of evidence theory are established with the cloud model algorithm of data driven. Examples validate that through fusion algorithm of the multifold fault components, the line selection is more robust and reliable to varied circumstance than utilizing individual component.

Tinnitus is an auditory phantom percept that affects the perception of sound in the patient’s ears, and the incidence of prolonged tinnitus is as high as ten to fifteen percent. Acupuncture is a unique treatment method in Chinese medicine, and it has great advantages in the treatment of tinnitus. However, tinnitus is a subjective symptom of patients, and there is currently no objective detection method to reflect the improvement effect of acupuncture on tinnitus. We used functional near-infrared spectroscopy (fNIRS) to explore the effect of acupuncture on the cerebral cortex of tinnitus patients. We collected the scores of the tinnitus disorder inventory (THI), tinnitus evaluation questionnaire (TEQ), hamilton anxiety scale (HAMA), and hamilton depression scale (HAMD) of eighteen subjects before and after acupuncture treatment, and the fNIRS signals of these subjects in sound-evoked activity before and after acupuncture treatment. According to the fNIRS detection results of tinnitus patients, acupuncture increased the concentration of oxygenated hemoglobin in the temporal lobe of tinnitus patients, and affected the activation of the auditory cortex. The study may reflect the neural mechanisms of acupuncture treatment for tinnitus and ultimately help to provide an objective evaluation method for the therapeutic effect of acupuncture treatment for tinnitus.

This paper proposes a hierarchical framework integrating deep reinforcement learning (DRL) and rule-based methods for multi-lane autonomous driving. We define an instantaneous desired speed (IDS) to mimic the common motivation for higher speed in different traffic situations as an intermediate action. High-level DRL is utilized to generate IDS directly, while the low-level rule-based policies including car following (CF) models and three-stage lane changing (LC) models are governed by the common goal of IDS. Not only the coupling between CF and LC behaviors is captured by the hierarchy, but also utilizing the benefits from both DRL and rule-based methods like more interpretability and learning ability. Owing to the decomposition and combination with rule-based models, traffic flow operations can be taken into account for individually controlled automated vehicles (AVs) with an extension of traffic flow adaptive (TFA) strategy through exposed critical parameters. A comprehensive evaluation for the proposed framework is conducted from both the individual and system perspective, comparing with a pure DRL model and widely used rule-based model IDM with MOBIL. The simulation results prove the effectiveness of the proposed framework.

The application of robots in high-precision automated machining is constrained by their limited multi-directional repeatability. In contrast, robots demonstrate superior levels of unidirectional repeatability, implying the potential for enhancing their precision. To further improve the positioning accuracy of robots, backlash error induced by rotating direction is considered, the concept of robot joint extended space is proposed, and the robot kinematic model is used to analyze the spatial similarity of robot error in the extended joint space. The dynamic Kriging method based on the optimization of basis functions is proposed to avoid overfitting of the surrogate model, and a model for estimating the robot’s positioning error in the joint extended space is constructed. Based on the estimated positioning error, the proposed calibration method is finally experimentally validated by error feedforward compensation. The results indicate that after Kriging interpolation in the robot joint space and feedforward compensation, the maximum/average positioning error of the robot is improved from 1.5157 mm and 0.8562 mm before compensation to 0.3471 mm and 0.1856 mm after compensation, and then further improved to 0.1848 mm and 0.1197 mm after adopting joint expansion space and dynamic Kriging interpolation, which decreases by 46.7% and 33.5%, respectively. This method effectively compensates the multi-directional repeatability error introduced by the joint backlash and improves the robot’s positioning accuracy.

In this paper, a microstructured laser with modulated periodic microstructures for beam control is proposed for the first time. Lasers using modulated periodic microstructures can emit beam with lower vertical divergence angle for the same number of microstructures. And lasers with specific period length can transmit beam in specific direction. These lasers emitting beams with different emission directions can be integrated into an array. And this array can be used directly as a transmitter chip of LiDAR. Both simulation and experimental results show that the modulated periodic microstructures can effectively reduce the vertical divergence angle. In the experiment, the lowest vertical divergence angle of 0.88° and a large angle sweep range of 41.5° have been obtained. A horizontal divergence angle of 4.3° is achieved with the taper structure. Typical optical power is 29 mW.

In this paper, a novel kind of GaN light-emitting diode (LED) with P and N discontinuous ohmic contact electrodes and discontinuous current blocking layer (CBL) is designed and prepared. Multiple contact windows under the N- and P-electrodes formed by etching help realize the discontinuous ohmic contacts with the N-GaN layer and the ITO layer, respectively. Similarly, by etching a plurality of discontinuous openings with the same pitch in the CBL, a discontinuous CBL structure is formed. As compared with traditional structures, the discontinuous ohmic contacts can maximize the number of current paths which allow the carriers to be collected from all four directions, leading to a more uniform current spreading and thermal management. Furthermore, as the contacts are not continuous any longer, some of the otherwise removed materials (such as the active region) lying between the patches are saved. The optoelectronic characteristics, junction temperature, thermal resistance, and infrared thermography measurements are conducted and the current distribution is simulated. The results show that the novel LED structure has better optoelectronic performances and current spreading ability compared to traditional devices. At 150 mA current, the luminous efficiency is 15.2% higher than the traditional LED, the junction temperature is reduced by 4.9%, and the thermal resistance is reduced by 12%.

Wireless sensor networks consist of a large number of nodes with constrained energy supply. Energy efficiency is hot and challenging issue in wireless sensor networks. Existing studies have shown that clustering is one of the efficient techniques to improve energy efficiency. An easy-to-implement and flexible method for even clustering and cluster head selection is beneficial to optimize network stability and energy efficiency. In this study, a sector-based lightweight and flexible clustering algorithm is developed to reduce node energy dissipation and optimize energy efficiency. Our proposal divides the area into virtual sectors. In the meanwhile, an even sector cluster is created using the sector decomposition approach. Based on the total communication distance, residual energy, and local node density, each node in the cluster calculates its own priority. The node with the highest priority is selected as the cluster head. Our proposal is compared with TSC,MH-TSC,SEECP,DREEP and LEACH. Experimental results show that the proposed algorithm outperforms these algorithms in terms of network stability and network lifetime.

The continuous development trend of expressways urgently needs expressway sensing, which is required to be able to perceive unknown risk factors to ensure a safe and stable transportation environment as well as effectively and continuously monitor the safe operation of expressways. Using unconscious payment can greatly improve traffic efficiency and meet the development direction of intelligent transportation. However, charging errors occur when using unconscious payment, primarily because the expressway network fails to realize national networking. Most of the literature focuses on two aspects of sensing, transmission and computing, while research on the overall network architecture is lacking. A ground network cannot solve the network coverage problem in remote areas. Therefore, the basic infrastructure needed to realize expressway sensing includes both aerial and ground nodes, such as cars, roadside equipment, unmanned aerial vehicles, airships, and satellites. This network communication architecture requires effective management of infrastructure and support of various applications. The software-defined network has strong flexibility; thus, in this paper, we propose a software-defined space-ground-vehicle integrated network (SD-SGVIN) architecture that includes three layers: physical facility, logic control, and application. The SD-SGVIN can separate the highway sensing applications from the physical entity. All kinds of applications can acquire, transmit, and process vehicle data through a shared physical infrastructure. We discuss the design, implementation, and advantages of the proposed network structure, hoping to provide a new direction for the research and promotion of expressway sensing and unconscious payment.

Quantum-dot color conversion film (QDCCF) with a partitioned black matrix (BM) realizes effective color generation with low crosstalk for micro-light-emitting diode (μLED) displays. However, this traditional BM seriously deteriorates light conversion efficiency (LCE) of QD converted μLED displays due to its absorption energy loss. In this paper, a novel internal and external reflective black matrix (IERBM) is proposed to improve the LCE and blue light utilization (BLU) by light recycling and reuse. Based on the IERBM structural optimization, the LCE can be respectively improved by 3.39× and 2.74× for red and green QDCC sub-pixels, and the BLU can be effectively increased by 1.11×. Accurate white balance and high color uniformity are achieved by this μLED display. Meanwhile, both the crosstalk of 1.72% and the color gamut of 89.8% Rec. 2020 are comparable with traditional absorptive BM. It is also found that the proposed IERBM has the potential to weaken μLED's sidewall effect. This work may open up a new technological route for improving the optical performance in advanced self-emissive displays.